Topline Impacts of My eCommerce Redesign

// Case Study Overview //

An impending end to federal Covid response healthcare subsidies predicted many people would soon lose healthcare coverage, therefore shop for it independently (not through an employer).

Kaiser Permanente (KP) recognized a timely opportunity to increase prospect traffic, conversion and retention in this segment and related ones.

In a highly competitive industry, KP's shopping experience had fallen behind. Previous modernization initiatives, though well intentioned, fell short.

Over half of shoppers had consistently reported feeling overwhelmed and confused, both by the actuarial jargon and the process as a whole. Despite feeling confused and overwhelmed, less than 2% of landing screen visitors used the "Get Help" feature, intended to guide shoppers to suitable coverage.

Those who did enroll in a plan also indicated low confidence that the plan they selected was the best fit for their needs, contributing to enrollment churn.

Business Context

I took a dual-track approach to research. On one track I first dove deep into KP's existing research, a goldmine of relevant studies in recent years to revive and augment with new formative research. Combined, they exposed an ongoing list of un-met shopper needs and wishes.

On the other track I began comprehensive competitive analysis. Auditing plan shopping experiences across major carriers and marketplaces to catalog which features, flows and information were common, vs unique.

Existing and new user research combined with my competitive analysis revealed which flows and features would reduce known pains, and improve behavioral metics.

I brought these research-based opportunity insights to our design sprint and facilitated participatory discovery and design sessions with executives and multiple cross-functional teams to align the organization behind a design approach to, by co-creating it together:

1. Reduce cognitive fatigue and confusion to increase conversion.

2. Increase confidence in plan selection to increase retention.

Initial Approach

"Know the user better than anyone" -Yours Truly

I transformed an intimidating, high-friction experience into a supportive, confidence-building one through user flow simplification and new features to present plan options in a more personalized context.

The old "Get Help" feature of last-resort was removed. Replaced by a simplified yet more comprehensive "Guide Me" flow: a question wizard with easy to answer questions.

I moved the plan search flow's entry point above the fold on the Shop Plans landing screen, to serve the low-knowledge majority who seek guidance. Instead of trial-and-error navigation. For those re-visiting at later journey stages, navigation to specific coverage types, based on age and eligibility, remained below, providing a section for each persona defined by journey stage.

I redesigned plan cards with color-coded metal tiers that can be recognized after coming from a government exchange site, because KP's UI is just one in a much larger ecosystem. And added contextual information to each plan about budget versus risk trade-offs of each plan, directly addressing reported "overwhelm," "confusion," and "low confidence."

Leveraging light weight automations further reduced user search friction: location auto-detection to load regionally available plans to start the search flow, and an AI model for personalized recommendations to completing.

Solution

Behavioral impacts

• Search flow time on task decreased 66% on average.

• Upper funnel lead capture increased 23%.

• Lower funnel bounce rate decreased 51%.

• "Help" vs "Guide Me" feature engagement increased 2,200%.

• "Help" vs "Guide Me" feature engagement conversion rate increased 4,875%.

Qualitative impacts

• Acceptance testing of the Question Wizard flow that replaced the old search form, scored 4.5 out of 5 across eight subjects.

Impacts

Back to top

Reflections

Shopping for healthcare differs significantly from retail eCommerce. The stakes are higher and it's more complex. It can involve your whole family, pricing has multiple facets for risk factor calculations, and enrollment normally requires a one year contract. KP's old design demanded that people submit most of their personal/family information before revealing any offers or value. Effectively asking them to invest time, energy and personal data in the process blindly—before any indication of value or affordability. Using a dating analogy, the old system played hard to get, revealing the offer only after shoppers jumped through a number of arduous hoops. It implied an asymmetrical relationship. I strove to flip that around by opening with a transparent offer of value up front, with plan previews. We could then refine plan options and costs as they add more personal context information. To establish an equal, reciprocal relationship that people would be more inclined to trust and commit to.

In practice, however, this change and others like it proved infeasible (backlogged) due to back-end and business rule complexities larger than our product team could unravel and improve. I left KP with a solid backlog of similar research-backed improvements to make, if/when possible. I hope they'll make some!